Thermal impulse from quenching red hot steel rod in water

"s.morra" wrote: (clip) I want to ask members of this newsgroup if you have experienced or know of any stories about a particular phenomenon when quenching a red hot steel rod in water. ^^^^^^^^^^^^^^^ You posted this same question a few days ago in rec.crafts.metalworking, and got eight responses, most of which totally discredited the supposed effect. Do you expect a different answer here? THERE IS NO SUCH THING AS THERMAL INDUCTANCE.

I have posted this question in several groups and I don't think there are any more pertinent groups left after this one (any suggestions?). Yes, I do expect and have seen different responses, as each group represents a different take on the world (diversity). Of course, I have also seen similar responses between the groups (steam on the hand, second law of thermo, doesn't exist, etc.), which I would expect.

I posted two basic questions, one regarding thermal inductance and the other regarding direct experience (or stories) with the phenomenon. You seem to be focusing on the thermal inductance part too much, which was my conjecture to explain the phenomenon and my first post (I thought my language was clear about conjecture, but maybe not). I won't argue about a conjecture and am glad others have debated it a bit, including you. Like you state, most don't think it is thermal inductance, you (below) go so far to state that there is no such thing, and I don't know. One reference I cited states otherwise (can't be dismissed) and the other doesn't dismiss it except to state the phenomenon hasn't been observed, and I only did a little bit of Google research. I am fundamentally interested in a viable explanation for the phenomenon (can't be ignored) and thermal inductance is only a guess (can be ignored). Steady state or quasi-static thermodynamics would be everyone's first thought on this problem, which leads to a second law of thermo problem (but read the cited reference). But there is an obvious highly non-linear process going on by quenching a red-hot rod in water and an obvious large thermal impulse occurring during the few second quench. Google research led to "extended irreversible thermodynamics".

You also stated in rec.crafts.metalworking that "I doubt that the fast heating of a quenched rod really happens". This is perhaps your belief or armchair opinion, as I appears you don't have any real experience with the phenomenon (do you?). However, there are now three people stating it does exist by their experience, that it is repeatable, that it has a pronounced effect to the hand, and is not due to steam on the hand (the palm contacting the metal gets hot, not the back of the hand). Discrediting is easy with no experience, something we all do. However after hearing the responses to my inquiry from several people in several groups, I decided to get a couple AD592 semiconductor temperature sensors and a USB computer DAQ module to take some measurements and investigate further (I'll post the data FYI). I got hold of a oxyacetylene torch but no tanks yet, so I might first try it using a blow torch and a smaller rod.

My inquiries into each group was to see if someone else has already made such measurements or new of some references about the phenomenon, prudent to save me time or from reinventing the wheel. Hopefully someone in this group (welding!) might have some direct experience with the phenomenon or could go out in their shop, try it a few times, and report back. Running an instrumented experiment only adds test data to a phenomenon that is plainly observed by one's hand. But some respondents feel this is important since they have never observed it nor have the means and I do appreciate and want their contributions towards a solution. However, such data might establish a few things like the propagation velocity (with sensors placed a foot apart along the rod), the sizes of the two impulses, their decay rates and shapes, etc. between a control and experimental rod.

Hi everyone

On this matter of "the heated rod, quenched or held in air":

The punch line is my bet would be on a perceptual thing like the sense of touch perceiving more temperature ("heat") in the held end of the rod, once away from the ambience of heat near the forge and now near the ambient coolness of the quench bath. Heat transfer in solids has long demonstrated itself to be simply behaved, so bets are not high on an explanation of heat really "racing away" up the rod away from the quench bath.

More detail:

Heat conduction in a solid like iron is a well-behaved (simply, orderly, regularly) physical process - about as regular as you meet in physics (?).

In my Doctoral work on hydrogen movement ("diffusion") in the weld zone of structural steel welds, heat flow was very much in immediate proximity of view, as heat *conduction* and mass-transfer *diffusion* follow the same mathematics. If there is a constant condition, like a steel plate 1inch thick, with boiling water on one side (100degC) and water with ice in it on the other side (0degC), after a while everything will settle down and there will be a constant flow of heat from hot to cold. In effect, the mathematics of more complicated situations simplifies-out and you get the expression for

*steady-state* heat transfer (or mass-transfer). In steady state heat transfer the property of the material (the iron) which affects rate at which heat passes through it is the *thermal conductivity*.

But that is not your situation. You are in the more general situation of changing conditions. This is known as *unsteady state heat conduction*. And *thermal conductivity* does not appear to the observer in itself. In unsteady-state conductive heat transfer you "observe" a material characteristic called the *thermal diffusivity". This is a composite the thermal conductivity and the heat capacity. Obviously, with a heat flux, some will "stay around" and some will "pass on by" (heat cap. and th. cond., resp.).

Units of thermal diffusivity, by the way, are "metres-squared per second".

One way to visual heat transfer is atoms "vibrating" in the hot part and "knocking against" neighbouring cooler atoms, passing on vibration and making these affected atoms "hotter". These in turn will "knock against" their neighbours; and so on, so on... Now don't use these visualisation like a crowbar to lever and attack at physicists talking about "phonons" and things like that - using another analogy - your "lever" has no "fulcrum"! But "vibrating atoms" gets you a long way. Certainly engineering qualification of heat transfer.

The reason for conjuring up this analogy is - these "vibrations" and "knocking against" have a certain "stiffness" and quenching in water a bar of iron does not do anything much to the "physical constants" you might derive. We are back to the "simply behaved" point.

The maths of heat flow will show that the held end of the bar will always be at a lower temperature at a given time when the hot end of the bar is quenched than it would be if the bar were "in air", cooling by convection of air and radiation to anywhere. That doesn't stop the held end of the bar getting hotter for a while, even after the hot end of the bar has gone into the quench water - with the heat "thermally diffusing" up there - but that temperature at time will be lower than for the "free" bar, as I said. I can know that because heat "flows downhill from hot to cold" in a simple regular way - it's just a simple relaxation downwards.

Hope this helps.

Richard Smith

My guess would be the "Leidenfrost Effect" This would insulate the hot end from the water which would force the heat up through the metal.

"JMV" wrote: My guess would be the "Leidenfrost Effect" (clip) ^^^^^^^^^^^^^^^ I had to look it up to be sure I knew what you were talking about. The Leidenfrost effect has to do with the difference between "film boiling" and "nucleate boiling." If the temperature difference between the water and the metal is above a certain figure, a film of steam forms an insulating barrier, so the boiling rate is low. This is called "film boiling." Below this temperature difference, the steam layer suddenly collapses, the water wets the metal, and the boiling rate goes up.

In order for this to produce the effect described by S. Morra, however, the heat transfer rate during film boiling would have to be lower that the heat transfer to still air. Film boiling produces steam, which requires much more heat than warming air in free convection.

Your idea of invoking the Liedenfrost effect was an interesting and worthwhile suggestion, but I don't think it holds up.

Wow, great guess! To me, this is the most plausible explanation of the dynamics I have heard so far, so hat's off to you JMV! See

for a description of the Leidenfrost Effect in lay terms. Whether this is the true and primary explanation or not for this phenomenon, the Leidenfrost Effect must be occurring in this case. As applied to the red hot rod (way over 200 C), the steam formed at the rod surface from plunging it into water must insulate the red hot region way more than normal as you note. So the red hot heat can't escape very well from the tip during the quench and perhaps more heat flows down the rod length somehow, which is the only other cooling path. If this were so and that was all there was to it, then the phenomenon could also be observed if the red hot end were inserted into an insulating sleeve, such as a vacuum thermos bottle (except the glass might break or melt with 1400 C nearby).

When I observed the phenomenon, it only took about 15 seconds or so before I could no longer hold the quenched rod (dropped it into the water) but I held the unquenched rod for a much longer time afterwards with no problem (only slightly warm). Although the insulation from the Leidenfrost Effect must be occurring, I still wonder if this was merely an increased heat conduction or there is a thermal wave, thermal inductive kick, or something else like that in operation. For both rods, the approximate thermal circuit is 1400 C at one end, 25 C at the other end, the thermal resistance and capacitance of the steel rod in between the ends, and the thermal resistance and thermal radiation into the air at the hot end (along the rod too, but this can be neglected compared to the hot end). Insulating the red hot end due to the Leidenfrost Effect doesn't change this circuit (hot end still at 1400 C, but dropping fast), excepting for greatly reducing the thermal resistance and thermal radiation into the air at the hot end. So I don't think the Leidenfrost Effect by itself really explains the strong thermal impulse felt by the hand from the quenched rod.

Good Day All

My first welding instructor tried to explain this phenomenon to us in overly simple terms; but it may be of use.

Both pieces of metal are loosing heat from the time they are removed from the fire. Steel is a poor conductor. The heat starts taveling along the bar at the same pace. When one bar is quenched the steam generated acts as an insulator between the bar and the water (quenching medium). Some of the escaping steam transfers heat to the bar as it escapes from the tank; this may only be an inch or so but apparently it makes enough of a difference that you will drop the rod being quenched first.

Apparently if the bar is inserted into a tank horizontally you don't experience this phenomenon; have never tried it to find out.

Hope this crude explanation is of some help.

John Noon

Let me pose a question then. In general, does the rod cool down quicker when quenched into water, or when in air? I believe that by a big margin the rod cools a slower in air, with or without the steam blanket effect. For sure, held steady in water, you can see the bar glowing red for some time - but that is still a faster cool that held in air. Regardless of the value of the discussion, I think this places a "bound" on the conjectures and says this, the "Leidenfrost Effect" brought into mention, cannot be the explanation.

??? Is this right?

I hadn't heard of this name before - the "Leidenfrost Effect". But I think the effect is very familiar, if I am making the correct association...

This effect is famous with water quench when you are trying to harden a steel. You want to cool fast in the range of about 900degC to around 550degC, to suppress transformation to soft products (ferrite and pearlite), then slowly from that temperature down to room temperature (limit stress differential during shear transformation to hard martensite). But with water you get the damn'd opposite. The vapour blanket makes the quench slow at high temperatures then with a great big "quack" sound the vapour blanket collapses and you get a fast final cool. It's so bad that a lot of quenching of hardenable steel is done into oil. I know - I used to have to wash the smokey oily residue off myself all the time when working in the test house of a steelworks heat treating samples.

Richard Smith

I would be curious to know if the heat transfer from the rod is greater or lesser with the water vapor blanket than in air. I would expect that the thermal resistance of the vapor layer is low, but the temperature is higher than air, so the differential driving heat from the rod to the surrounding gas is lower in the vapor than in air....

This question posed (and I really am curious), I tried this today, and didn't notice any difference quenching the tip in water than just holding the rod. Test conditions: rods at same heat color, soaked in fire equal time. Held using thin leather gloves (TIG gloves). Rods kept at (roughly) same orientation in each hand as one was quenched. Tried it a couple times, quenching with alternate hands. The only thing I noticed was that I tended to grip the rod I was going to quench more tightly--not for any concious reason, maybe because I was more concerned with where it was going?

Even if the steam boundy layer reduces heat loss from the rod tip, I very much doubt that it would cause the claimed effect of rapid heating at the other end of a three foot rod. I really think there must be some perceptual effect, or possibly the (hot) water vapor (which is not visible, as it is a gas) bringing heat to the glove, and possibly condensing, carrying the heat more readily through the glove.

wrote: (clip) I tried this today, and didn't notice any difference quenching the tip in water than just holding the rod.(clip) ^^^^^^^^^^^^^^^ What do you know? An actual experiment! Yet, I'll bet it doesn't end the discussion. The OP has already been to several forums, and if the outcome here doesn't please him, he'll find another place, and start all over.

enl_public, thanks for trying to experience the phenomenon. If you have the chance to try again, I'd suggest removing the gloves and using bare hands. I don't think it is a strong enough phenomenon to clearly detect by the hand through gloves.

To Mr. Lichtman, take a powder and chill out. Recapping, "What do you know? An actual experiment! Yet, I'll bet it doesn't end the discussion. The OP has already been to several forums, and if the outcome here doesn't please him, he'll find another place, and start all over." I'm out of places, don't intend to repeat myself, a reasonable man, and seeking an answer from newsgroups' collective wisdom. A Google search on you still shows you to be similar, not withstanding your apparently hostile and undeserved projection here. Killfile the thread if its that troubling to you.

I'm delighted for an independent confirmation (or not), especially by anyone who welds. It appears that enl_public set up and executed the experiment correctly, excepting for wearing gloves (obvious problem). Regardless, I will set up an analytic experiment and post the data on one of the binaries newsgroups of the outcome. I have the temperature sensors and should receive the DAQ module tomorrow.

I've experienced this many times. The guy (now departed) that taught me much about welding and metal things in general, always used to call it 'chasing'. When you quenched a piece of metal, he always said it 'chased' the heat up. I could never figure out what was going on though. The steam may be the culprit. You may want to try shielding the top of the piece from the steam when you do your tests.

Bob

Well, it's been nine years, is anyone in this thread still alive and/or around? I'm curious about this as well, as I've experienced something like this many times -- with Magnalite cookware.

[IMG] I have a large Magnalite Professional (magnesium steel alloy) skillet. Attached to it is a long thick aluminum handle, which is always at thermal equilibrium with the pan (~375-400 deg. F) prior to quenching. Covering the handle is a silicon "Stay Cool" insulating cover, which still gets fairly hot during usage but you can at least touch it. When I'm finished with the hot pan I sometimes quench the underside rapidly under cold running water (horizontally, so zero steam is involved with the handle). When I do that the covered part of the aluminum handle becomes insanely hot to the touch.

My conclusion is that heat is being "chased" up the aluminum handle during quenching. Physically it makes sense as a transient phenomenon. As rapidly vibrating (hot) molecules at the quenching boundary layer impart some of their kinetic energy to the cold water, an equal reaction in the opposite direction takes place as energy also rebounds, propagating like a wave in the direction of the handle. That kinetic energy is then imparted (added) to the already hot molecules in the handle, effectively condensing the heat there in what is a transient process.

This is patent nonsense.

i

An old blacksmith technique for tempering a long slender object - think boring bar - was to heat to "cherry-red" and quench the end in come medium. Then, quickly, hit it with a grind stone or even rub it on the cement floor, so that you have clean bright steel and then watch the tempering colors run down the shank until the cutting edge turns the correct color and put it back in the quench medium.

So, in that instance one might say that the heat didn't run up, it ran down :-)

You just can't see the steam.

It resembles the old Phlogiston arguments.

It 'resembles' phlogiston arguments? I'm actually not arguing, so much as questioning to discover whether or not anyone has finally put this to a quantitative test -- not to read any irrelevant dismissive non-arguments from ignorance. If you're going to posture yourself as being on the side of real science, at least be somewhat scientific about it.

Perhaps I couldn't "see the steam" (outside the copious amounts of steam I could actually see during quenching, far from the handle), but that should change soon enough. I've ordered a FLIR thermal imaging camera to quantitatively capture transient responses throughout the whole of the pan, which will hopefully give an actual meaningful answer one way or another.

The test is simple. I'll begin with a handle that is fully shielded and insulated from any possible interaction with any steam. If that handle does not get measurably hotter during quenching, then I will conclude that my senses have fooled me and that the phenomenon is not real. If the handle does get measurably, temporarily hotter, then I think it will be safe to conclude that the phenomenon is real. It's that simple, having nothing to do with anyone's opinions or ad hoc explanations. If you can think of a better test or see a flaw in my methodology, by all means chime in.

Conductive heat flow is simply "downhill" from hot to cold. No matter how big the colder reservoir and how "tempting" the hotter part might try to be in some way.

Nothing complicated like fluid flow (?). I can write computer programs to calculate conductive heat flow in

3-dimensional objects, but I can hardly start to begin to understand fluid flow. Heat doesn't do anything fancy like has been described.

The one about quenching one end of a long tool, quickly buffing it to silver then letting back-flow of heat take it through a tempering-temperature-cycle would be for real. I've heard about a similar one for making high-strength high-properties "rebar" (reinforcing bar for concrete) - controlled quench then withdraw quench when there is enough heat in the bar's core to "self-temper" the quenched structure. For sure you get a varying structure between core and surface, but it's a lot of strength cheaply and every property met - it's "fit-for-purpose".

etc.

Rich